Skip to main content

We are quants.

What guides us

Some investors are guided by fundamentals, others by macroeconomic, structural or technical indications. We use quantitative methods and machine learning for a holistic view of styles to identify which style works, at which time, in which market.

We are driven to capture the highest and most robust returns possible, across all cycles and regimes. Robustness always comes before performance. We  detect market dogma and take advantage of the delta that exists between our models and market opinion.

Signal building

Each of our alpha signals must compete against existing research and a rigorous testing cycle. Only when we can identify an advantage does the signal find its way into the portfolio.

Sub-strategy portfolios

We think in terms of a variety of strategies, which can exploiting different investment styles. Since each of these signals has its own specific characteristics, we aggregate the signals across many assets into a sub-strategy.

Meta portfolio

Our meta portfolio is aggregated based on the movement patterns of the sub-strategy portfolios. The meta portfolio is constantly reassembled in order to optimally exploit the different behaviors of the sub-strategies.

Trading cost reduction

The use of various statistical tools and machine learning, allow us to implement strategies even when high turnover occurs. We also view cost reduction as an alpha component.

We are asset class agnostic. Furthermore, we classify each tradable instrument based on its characteristics and consider it with similar assets in a group. By constantly revaluing asset groups, we can reliably anticipate structural changes. In this way, we maintain maximum diversification potential at all times.

Our quality assurance is at the core of our beliefs. Every assumption, every signal and every portfolio, must stand up to a series of validation mechanisms that include the largest possible number of historical regimes and structural breaks. To test the unprecedented as well, we synthetically generate time series that contain as yet unknown but imaginable regimes and correlations. In this way, we create models that can be trusted.

Who we are

We are committed data scientists with a drive to explore new things and improve old ones. We seek to understand what the meta levels and first principles of our models, assumptions, and ideas are. Our different schools of thought allow us to constantly question the status quo. As an interdisciplinary team with experience in managing multi-billion EUR AUM portfolios, data science and software development, we provide exceptional returns.

Lukas Widmann

Lukas formerly worked for one of the largest German institutional investors as a portfolio manager and is co-founder of an AI start-up focusing on time series analysis. For its models, dcorr has already been declared a winner at the Future Fundstars Award (Fundview). dcorr draws on a broad network of practitioners in the financial and AI industry to offer best-in-class performance over the long term.


We look forward to hearing from you.